As you noted, the budget was only approved at full amount once in 20 years. So, what probability would you assign to this year's budget being approved at the full amount?
By the way, this is a very good example of why you need to apply statistics very carefully. Surely, a straight assignment of this year's probabilities based on a 20 year history is not going to be accurate. Perhaps it's the best one can do, but I doubt that. An expert can assign probabilities based on current information much better in this case.
For example, most likely the year that the budget was fully approved was a year when there was a strong drive among politicians and citizens to spend money on military. If the policitical climate has changed, an expert may well know that the chance of approving a full budget is absolutely zero. In such a case, using a 5% probability based on past history is ridiculous.